Get most probable words for each topic

Question

I made an LDA model with sklearn but, as weird as it sounds, I cannot find anything online about how to get the top-words. This is my code:

from